General Reasoning on Overall MATH-500 AIME25 HumanEval GPQA
85.1AccuracyVanilla
Evaluation results
| Method | Links | ||
|---|---|---|---|
| VanillaBackbone=Ministral-3-8B-Reasoning-25122026.03 | 85.1 | 100 | |
| VanillaBackbone=Ministral-3-14B-Reasoning-25122026.03 | 83.4 | 100 | |
| DynasorBackbone=Ministral-3-8B-Reasoning-25122026.03 | 79.9 | 73.9 | |
| TERMINATORBackbone=Qwen3-14B2026.03 | 78 | 65 | |
| VanillaBackbone=Qwen3-8B2026.03 | 77.4 | 100 | |
| VanillaBackbone=Qwen3-14B2026.03 | 77 | 100 | |
| DynasorBackbone=Ministral-3-14B-Reasoning-25122026.03 | 76.9 | 75.1 | |
| TERMINATORBackbone=Ministral-3-14B-Reasoning-25122026.03 | 76.5 | 65 | |
| TERMINATORBackbone=Ministral-3-8B-Reasoning-25122026.03 | 73.4 | 64.7 | |
| TERMINATORBackbone=Qwen3-8B2026.03 | 72.6 | 67.8 | |
| Thought-CalibBackbone=Qwen3-14B2026.03 | 70.6 | 83.1 | |
| Thought-CalibBackbone=Qwen3-8B2026.03 | 70.1 | 86.8 | |
| DEERBackbone=Ministral-3-8B-Reasoning-25122026.03 | 68.8 | 78.9 | |
| DynasorBackbone=Qwen3-14B2026.03 | 67.7 | 51.6 | |
| Thought-CalibBackbone=Ministral-3-8B-Reasoning-25122026.03 | 66.5 | 84.5 | |
| DynasorBackbone=Qwen3-8B2026.03 | 62.2 | 49.2 | |
| DEERBackbone=Qwen3-14B2026.03 | 59.6 | 78 | |
| NoThinkingBackbone=Ministral-3-8B-Reasoning-25122026.03 | 59.3 | 14.6 | |
| DEERBackbone=Qwen3-8B2026.03 | 57.3 | 75.8 | |
| Thought-CalibBackbone=Ministral-3-14B-Reasoning-25122026.03 | 56 | 86.5 | |
| NoThinkingBackbone=Qwen3-14B2026.03 | 55.2 | 17.1 | |
| NoThinkingBackbone=Ministral-3-14B-Reasoning-25122026.03 | 55.2 | 13.7 | |
| NoThinkingBackbone=Qwen3-8B2026.03 | 53 | 15.6 | |
| DEERBackbone=Ministral-3-14B-Reasoning-25122026.03 | 50.9 | 75.3 |